Management of innovation processes

Good summary for this exam !

Onderzoeksvaardigheden 1 IW (Universiteit Utrecht)

Verspreiden niet toegestaan | Gedownload door Ben De Groot ([email protected])

lOMoARcPSD

http://www.studeersnel.nl/nl

http://www.studeersnel.nl/nl/document/universiteit-utrecht/onderzoeksvaardigheden-1-iw/summaries/good-summary-for-this-exam/25905?auth=1&auth_prem=1&prem_doc=1&testVersion=

http://www.studeersnel.nl/nl/course/universiteit-utrecht/onderzoeksvaardigheden-1-iw/4467?auth=1&auth_prem=1&em=0

H1 Socail research strategies

Research is done in order to give answers to questions. It can be done to enlarge general knowledge

or to solve some sort of problem. Theories are used in research as analytical tools for understanding

and explaining a subject matter. Theory has logical non-contradictory propositions, is testable and

has an explanatory mechanism.

Theory can be divided into different levels. They are stated below with their (dis)advantages:

- Grand theories: abstract, difficult to use

- Middle range theories: focuses on limited aspect, easy to use

- Empirical findings: observations

There are different ways to perform research, but can also be used together:

- Deduction: Theory testing by doing observations, quantitative, trial and error

- Induction: Theory making out of observations, qualitative, finding a pattern

Epistemology is about what is considered acceptable knowledge:

- Positivism: methods of the natural sciences should be used to study social reality, objective

observations, there is one absolute truth, deduction

- Realism: various forms of realism exist, they are in between positivism and interpretivism

- Interpretivism: interpretation of research plays major role, subjective meaning, there are multiple

truths, induction.

Ontology is concerned with the question what social entities are:

- Objectivism: social phenomena exist independent from social actors.

- Constructionism (constructivism): social phenomena are constructed by social actors, social

interaction ensures continuous revision, multiple interpretations of researcher

Values: Personal beliefs or feelings of

researcher, affect every stage of

research process

Practical considerations: possible

constraints; time; cost; existing

literature etc., compromise between

ideal and feasible.

Research strategies are twofold:

quantitative and qualitative.

H2 Research design

There are several important criteria to evaluate social research. The most important are stated and

explained below:

- Reliability: whether the measurements performed are consistent

- Replication: whether the study is replicable

- Internal validity: whether the measurements represent the concepts

Quantitative Qualitative Goal Measurement of social variables Understanding subjective meaning Role of theory Deductive Inductive Epistemology Positivism or realism Interpretivism Ontology Objectivism Constructionism Data Statistical and numerical Words, texts, stories Research design Surveys and experiments Interviews and ethnography


lOMoARcPSD


- External validity: whether the results can be generalized

A research design is a framework for the collection and analysis of data. The research method is the

way you collect the data (interview, observation etc.). Five main research designs:

- Experimental: establish causal relationship between independent and dependent variables, only

design with manipulation and control groups, rarely used in innovation research: impractical or

unethical

- Cross-sectional: collection of data from more than one case at a single point in time, shows variation

between subjects quantitative

- Longitudinal: survey of same sample at different points in time, shows change over time, quantitative

- Case study: detailed and intensive analyses of one case, qualitative, problems: hard to be objective

and difficult to generalize.

- Comparative: using same methods to compare two or more contrasting cases, problems: finding

comparable samples

Summary 3 OV 1

Planning a research project

- Follow always the requirements, instructions and information you institution has given to you

- Start early with thinking on research questions.

- Use your supervisor to the fullest extent that you are allowed and follow the pointers you are given

by him or her.

• Respond positively on his critics.

• If you get stuck or get behind with your work don’t avoid your supervisor � you will get in a negative

vicious circle.

Managing time and resources

- Work out a timetable

- Find out what resources can be put at your disposal for carrying out your research.

Formulating suitable research questions

- Qualitative research � more open-ended than

- Quantitative research

• very open-ended research is risky � collection of too much data

- No research questions or poor ones lead to poor research.

- If you do not specify clear research questions, there is a great risk that your research will be

unfocused and that you will be unsure about what your research is about.

- You must be clear about your research questions, they are crucial because they will:

• guide your literature search

• guide you in deciding what data you need to collect

• guide your analysis of your data

• guide your writing-up of your data

• stop you from going off in unnecessary directions and tangens

- Research questions in quantitative research are sometimes more specific than in qualitative

research. Some qualitative researchers advocate a very open approach wit no research questions.

Although this is very risky

- When we begin a research we usually start with a general research area. That may derive from:

• Personal interest/experience

• Theory

• The research literature


lOMoARcPSD

• Puzzles

• New developments in society

• Social problem

- The research area has to be narrowed down so that we can develop a tighter focus out of which

research questions can be developed.

- Steps in selecting research questions: Research area � Select aspect of research area � Research

questions � Select research questions

- We can’t answer all the research questions that occur to us, we select from the possible questions

that we arrive at. We should be quided by the principle that the research questions we choose

should be related to one another.

- Things that are important while writing a research proposal:

• What is your research topic, or alternatively, what are your research objectives?

• Why is your research topic (or why are those research objectives important)

• What is your research question or what are your research questions?

• What does the literature have to say about your research topic/objectives and research questions?

• How are you going to go about collecting data relevant to your research questions? In other words,

what research methods are you intending to use?

• Why are the research methods/sources you have selected the appropriate ones for your research

questions?

• What resources will you need to conduct your research and how will those recources be funded?

• What is your timetable for the different stages of the project?

• What problems do you anticipate in doing the research?

• What are the possible ethical problems associated with your research?

• How will you analyse your data?

Prepering for you research:

- Do not begin with data collection until you have identified your research questions reasonably

clearly.

- Some other questions you have to ask before doing research:

• Who do you need to study in order to investigate your research questions?

• How easily can you gain acces to sampling frame?

• What kind of sampling strategy will you employ

• Can you justify your choice of sampling method?

Hints about doing research

- Keep good records of what you do

- Make sure you are familiar with the hardware that you use for you research.

- Do not wait until all your data have been collected to begin coding.

- Remember that the transcription of recorded interviews takes a long time.

- Become familiar with any data analysis packages as soon as possible

- Do not at any time take risks with your personal safety.


lOMoARcPSD


Summary 4 OV 1

The purpose of exploring the existing literature should be to identify the following issues:

• What is already known about this area?

• What concepts and theories are relevant to this area?

• What research methods and research strategies have been employed in studying this area?

• Are there any significant controversies?

• Are there any inconsistencies in findings relating to this area?

• Are there any unanswered research questions in this area?

- During your reading you have get the most out of it.

• Make good notes

• Develop critical reading skills

• Use your review of the literature as a means of showing why your research questions are important.

• Do not try to get everything you read in your literature review.

• Don’t stop with reading literature once you begin research

- Systematic review: An approach to reviewing the literature that adopts explicit procedures. It has

emerged as a focus of interest because:

• Many reviews of literature tend to ‘lack thoroughness’ and reflect the biases of the researcher

• Systematic reviews are often seen as an accompainiment to evidence-based approaches.

- Some of the main steps of systematic review:

• 1. The purpose of the review must be definend.

• 2. Establish criteria to guide the selection of studies.

• 3. The reviewer should seek out and incorporate within the review of all studies that meet the

criteria of step 1. The search will be base don keywords and terms relevant to the purpose defined in

step 1.

• 4. Key features of each study should be identified. A formal protocol should be used to record

features: date when the research was conducted; location.

• 5. A synthesis of the results then has to be produced. Like; summary statistics from the quantitative

data.

- Limits of Systematic review:

• Situations where research questions are not capable of being defined in terms of the effect of a

particular variable or when the subject boundaries are more fluid and open or subject to change.

• It can lead to a bureaucratization of the process of reviewing the literature.

• Methodological judgments inform decisions about quality and so determine the inclusion or

exclusion of an article from a literature view.

- Wheter or not the systematic review makes sense to you depends on you own knowledge of the

theory. (epistemological position)

- Narrative review: Purpose is to enrich human discource by generating understanding rather than

by accumulating knowledge.

• It is a more uncertain process of discovery.

• Narrative reviews tend to be less focused and more wide-ranging in scope than systematic reviews.

• Less explcit about the criteria for exclusion or inclusion of studies

- Narrtive review is more suitable for qualitative research.

- Most reviews are of the narrative kind.

- Main methods of referencing your work:

• Harvard or author-date: if you are referencing to an author you put immediately after the argument

the name and the year of publication. If you are quoting you put quatation marks around the

quotation and after the year of publication. You include the page number where the quatation is

from.


lOMoARcPSD

• Footnote or numeric: This approach involves the use of superscript numbers in the text that refer to

a note at the foot of the page or the end of the next.

- A good bibliography should reflect the author’s informed judgement.

- Several difficulties why plagiarism is an issue:

• Universities vary in their definition of what plagiarism is.

• They vary in their response and punishment when it is recovered.

- Internet main source behind plagiarism.

5&6 Stances on ethics

Authors on social research ethics can be characterized in terms of the stances they take on the issue.

The following stances can be distinguished.

Universalism. Ethical precepts should never be broken. Infractions of ethical principles are wrong in a

moral sense and are damaging to social research.

Situation ethics: - the end justifies the means. Unless there is some breaking of ethical rules,

we would never know about certain social phenomena.

- no choice. It is often suggested that we have no choice but to engage in dissimulation on occasions if

we want to investigate the issues in which we are interested.

Ethical transgression is pervasive. It is often observed that virtually all research involves elements

that are at least ethically questionable. This occurs whenever participants are not given absolutely all

the details on a piece of research, or when there is variation in the amount of knowledge about

research.

Anything goes (more or less). Very few researchers subscribe to this stance. Denzin (1968) comes

close to an anything-goes stance when he suggests that social researchers are entitled to study

anyone in any setting provided the work has a ‘scientific’ purpose, does not harm participants, and

does not deliberately damage the discipline.

Deontological versus consequentialist ethics. Deontological ethics considers certain acts as wrong (or

good) in and of themselves. Consequentialist ethics looks at the consequences of an act for guidance

of whether it is right or wrong.

Ethical principles

Four main areas of ethical principles by Diener and Crandall (1978):

- whether there is harm to participants

The ASA code of ethics suggests that, if there is any prospects of harm to participants, informed

consent, the focus of the next section, is essential: ‘informed consent must be obtained when the

risks of research are greater than the risks of everyday life. Where modest risk or harm is anticipated

informed consent must be obtained.

- whether there is a lack of informed consent


lOMoARcPSD


covert observation for example, transgresses that principle, because participants are not given the

opportunity to refuse to cooperate. They are involved, whether they like it or not.

- whether there is an invasion of privacy

Personal information concerning research participants should be kept condifential. In some cases it

may be necessary to decide whether it is proper or appropriate to record certain kinds of sensitive

information.

- whether deception is involved

- deception occurs when researchers represent their work as something other than what it is.

- Ethical research is of a high quality. Thus, if a study is poorly designed, quite aside from the fact that

it almost certainly would not receive financial support from the ESRC (Economic and Social Research

Council), it is unethical.

- Research staff and subjects must be informed fully about the purpose, methods and intended

possible uses of the research, what their participation entails and what risks, if any, are involved.

- The independence of research must be made clear, and any conflicts of interest or partially must be

explicit. This draws attention to the possible role of affiliation bias to which some writers on ethics in

research draw attention (Bell and Bryman 2007).

Voor de overige 3 koppen,

ethics and the issue of quality:

ethical issues sometimes become difficult to distinguish from ones to do with the quality of research.

Discussie wanneer kwaliteit van het onderzoek belangrijker is of wanneer je rekening moet houden

met de ethische waarden en je kwaliteit moet inleveren.

the difficulties of ethical decision-making:

the internet and other new media have opened up new arenas for ethical decision-making.

Politics in social research:

There are political dimensions to the research process that link with the place of values.

The political dimensions of research are concerned with issues to do with the role and exercise of

power at the different stages of an investigation.

The nature of quantitative research


lOMoARcPSD

Research is rarely as linear and as straightforward as the figure implies, but its aim is to do no more

than capture the main steps and to provide a rough indication of their interconnections. It represents

a useful starting point for getting to grips with the main ingredients of the approach and the links

between them.

Concepts are the building blocks of theory and represent the points around which social research is

conducted (like structure, culture, social class, etc.)

Once concepts are measured, concepts can be in de form of independent or dependent variables.

Independent: an explanation of certain aspects of the social world

Dependent: things we want to explain

Why measure?

1. Measurement allows us to delineate fine differences between people in terms of he characteristic in

question.


lOMoARcPSD


2. Measurement gives us a consistent device or yardstick for making such distinctions. (Consistent over

time and other researchers)

3. Measurement provides the basis for more precise estimates of the degree of relationship between

concepts (like correlation).

In order to provide a measure of a concept, it is necessary to have an indicator that will stand in for

the concept.

Ways in which indicators can be devised (and how obtained):

- Through a question (structured interview schedule/self-completion questionnaire)

- Through the recording of individual behaviour (observation schedule)

- Through official statistics (spreekt voor zich)

- Through an examination of mass media content (content analysis)

Likert scale:

The goal of the Likert scale is to measure intensity of feelings about the area in question.

Why use multiple-indicator measures?

- It is possible that a single indicator will incorrectly classify many individuals.

- One indicator may capture only a portion of the underlying concept or be too general.

- You can make much finer distinctions.

A concept comprises different dimensions. When the researcher is seeking to develop a measure of a

concept, the different aspects or components of that concept should be considered.

BUT it would be a mistake to believe that investigations that use a single indicator of core concepts

are somehow deficient.

What is crucial: whether measures are reliable and whether they are valid representations of the

concepts they are supposed to be tapping.

Reliability is fundamentally concerned with issues of consistency (samenhang) of measures.

Factors whether a measure is reliable:

- Stability. Test-retest method: high correlation � reliable

- Internal reliability. We need to be sure that all our designerism indicators are related to each other.

One way of testing internal reliability is the split-half method. The degree of correlation between

scores on two halves of indicators can be calculated.

- Inter-observer consistency. When a great deal of subjective judgement is involved and where more

than one observer is involved, there is possibility that there is al lack of consistency in their decisions.

Validity = whether a measure of concept really measures that concept.

Forms of validity:

- Face validity. The measure apparently reflects the content of the concept in question.

- Concurrent validity. The researcher employs a criterion on which cases are known to differ and that

is relevant to the concept in question.

- Predictive validity. The researcher uses a future criterion measure, rather than a contemporary one

(as in the case of concurrent validity).

- Construct validity. The researcher is encouraged to deduce hypotheses from a theory that is

relevant to the concept.


lOMoARcPSD

- Convergent validity. Comparing a measure to other measures of the same concept developed

through other methods.

Reliability and validity are analytically distinguishable, but they are related because validity presumes

reliability. If a measure is not reliable, it’s not valid.

4 main preoccupations of quantitative researchers:

- Measurement. (zie ‘ Why measure’ )

- Causality. Quantitative researchers are rarely concerned merely to describe how things are, but are

keen to say why things are the way they are. An experimental design is more highly valued than a

cross-sectional research, because of the greater confidence that can be enjoyed in the causal findings

associated with the former.

- Generalization. In quantitative research the researcher is usually concerned to be able to say that

his or her findings can be generalized beyond the confines (grenzen) of the particular context in

which the research was conducted.

- Replication. As a check upon the influence of biases and lack of objectivity, scientists may seek to

replicate (reproduce) each other’s experiments. If there was a failure to replicate, so that a scientist’s

findings repeatedly could not be reproduced, serious questions would be raised about the validity of

his findings.

Chapter 7 Sampling

Selecting individuals for survey research

The researcher has to decide what kind of population is suited to the investigation of the topic and

needs to formulate a research instrument(structured interview schedule or a self-completion

questionnaire). In quantitative research sampling is always needed.

Important with sampling is being able to generalize findings, so they represent the whole population.

In order to generalize, the sample must be representative.

Basic terms and concepts in sampling:

Population: the universe of units from which the sample must be selected.

Sample: segment of population that is selected for investigation.

Sample size: depends on a number of consideration, no definitive answer:

Absolute sample size: sample of 1000 in US has as much validity as a sample of 1000 in USA. Increasing size

increases the precision of a sample and sampling error decreases.

Costs: larger size of sample mean comes converges more to population mean, but the convergence occurs at a

decelerating rate. (very large samples are decreasingly cost efficient).

Non-response: use methods to keep it as low as possible.

Heterogeneity of population: greater heterogeneity->larger sample needed.

Kind of analysis: some kinds need a very big sample.

Sampling frame: listing of all units in the population from which the sample will be selected.

Representative sample: sample that reflects the population accurately.

Sampling bias: distortion in the representativeness.

-when some of the sampling frame stand little or no change of being selected for inclusion in the

sample.

-Sampling frame is inadequate.

-Non response


lOMoARcPSD


Probability sample: method of sampling, using random selection so every unit has a known chance

of being selected. (more representative, less sampling error).

-Simple random sample: giving each unit in the population a numbers and let computer generate random

numbers) Sampling fraction: n/N n= sample size N=population size.

-Systematic sample: Variation on simple random sample: When 1 of 20 of population is chosen, number

between 1 and 20 is picked and then every time plus twenty. (16->36,56,76). There must be no ordering in the

sampling frame!

-Stratified random sampling: stratifying the population by a criterion. The resulting sample will be distributed

in the same way as the population in terms of the stratifying criterion.

-Multi-stage cluster sampling:

Non probability sample: some units have higher chance to be selected.

-Convenience sampling: sample that is simply available to the researcher by virtue of its accessibility.

-Snowball sampling: Researcher makes initial contact with small group and uses these to establish contacts

with others.

-Quota sampling: for commercial research, market research and political opinion polling.

Produce a sample which reflects the population in terms of the relative proportions of people in different

categories.(gender, ethnicity, age groups or combinations). Unlike stratified sample, the sampling is not carried

out randomly, final selection is left to the interviewer. By using census data it is known how many people in

each group should be used to reflect the population.

Sampling error: error in findings due to difference between sample and population, can occur even

when probability sample has been used.

Non-sampling error: error in findings that occur by deficiencies in the sampling approach(inadequate

sampling or non-response) or by problems as poor question wording, poor interviewing, or flawed

processing of data.

Non-response: members of the sample refuse to cooperate, cannot be contacted or for some reason

cannot supply the required data.

Census: enumeration(establish the number) of an entire population.

Hoofdstuk 8

The structured interview

Questions are usually very specific and very often offer the interviewee a fixed range of answers (

closed, closed ended, pre-coded or fixed choice). The goal of this sytle is to ensure that interviewees’

replies can be aggregated and this can be achieved reliably only if those replies are in response to

identical cues.

Common sources of error:

- A poorly worded question

- the way the question is asked by the interviewer

- misunderstanding on the part of the interviewee

- memory problems on the part of the interviewee

- the way the information is recorded by the interviewer

- the way the information is processed, either when answers are coded or when data are entered into

the computer

variation = true variation + variation due to error; aim is to keep the error to a minimum . error has

an adverse effect on the validity of a measure.

Types of interviews: structured or standardized interview, (semi-structured, unstructured(only list of

topics) or intensive, in-depth)�general term is qualitative, focused and Group interview.


lOMoARcPSD

There are more interview contextst than the face-to-face one, like “more than one interviewee”,

“more than one intervieuwer” or by telephone.

Advantages disadvantages telephone……………………………….?????

Conducting interviews: You have to know the schedule, introduce your research properly, achieve a

quick relationship with the intervieuwee, Ask you questions, make clear they are recorded, give clear

instructions and think about the question order (important).

Problems with structured interviewing (Ik verwacht hierover niet sin tentamen)

- characteristics of interviewers

-Response sets

- the problem of meaning

-feminist critique

Algeheel denk ik dat dit hoofdstuk weinig toevoegt want wij hebben dit gedaan in een survey en dat

is volgens mij ook volgens frank, beter! Want in dit hoofdstuk gaan ze ervan uit dat je deze ‘enquete’

form uitvoord als interviewer-interviewiee . Anton, je hebt je best gedaan, we zijn tevreden met je. Is

uitvoord trouwens een oosters woord?

H9&H10

Welcome to the fantastic short version of Chapter 9 and chapter 10 from Alan Bryman’s book on

social research methods. These chapters are found in part two of the book, which is about

quantitative research. Gemaakt door: Marcol dol koochol

We will start with chapter 9, which is called ‘Self-completion questionnaires’. Basically, this is the

‘enquete’ which you used in your research.

Introduction

Most of the information of the previous chapter also applies on self-completion questionnaires.

Evaluating the self-completion questionnaire in relation to the structured interview

In a SCQ (= Self-Completion Questionnaire), there is no interviewer present.

Further, - there are fewer open questions

- they have easy to follow designs

- they are shorter

Advantages of the self-completion questionnaire over the structured interview:

- they are cheaper to administer

- they are quicker to administer

- there is no interviewer present, so the interviewer does not influence the respondent (absence of

interviewer effects)

- there is no interviewer variability (when there are different interviewers in a research, their

interpretation of the answers respondent give may differ, this is called interviewer variability)

- it is convenient for respondents

Disadvantages of the self-completion questionnaire in comparison with the structured interview:

- the researchers cannot prompt (prompting is helping someone who doesn’t understand the

question)

- the researchers cannot probe (probing is elaborating on answer, doorvragen)

- you cannot ask many questions that are not salient (interesting) to respondents

- you cannot ask to many open questions


lOMoARcPSD


- the questionnaire can be read as a whole, so respondents can read the last question first etc.

question order may not be used by the respondents

- you cannot be sure that the right person answers (for example, you receive a questionnaire by mail,

your sister fills it in and sends it back)

- you cannot collect additional data (makes no sense to me)

- you cannot ask a lot of questions, because of respondent fatigue

- a SCQ is not suited for every respondent (for example, an English SCQ for someone who doesn’t

speak English)

- there is a greater risk of missing data

- there are lower response rates

Steps to improve response rates to postal questionnaires:

- add a good introduction

- add stamps to return the questionnaire

- send it again

- keep it short

- make a nice layout + clear instructions

- start with interesting questions

- personalize it, add someone’s name and address

- as few open questions as possible

- give them money!

Designing the self-completion questionnaire

- do not cramp the presentation (text to close to each other)

- make a clear presentation, nice layout etc.

- vertical closed answers are preferred over horizontal closed answers, this is more clear for the

respondent and easier to code.

- Tip van Frank: switch between positively asked questions and negatively asked questions to pick out

the people who just randomly pick an answer (acquiescence)

- give clear instructions on how to respond

- keep questions and answers together, don’t spread them over two pages

Diaries as a form of self-completion questionnaire

3 different versions:

1) a scheme constructed by the researcher that should be filled in by the respondent

2) a document written by the respondent itself

3) a log

Advantages of the diary as a method of data collection:

- it is more valid, and more reliable than a questionnaire

- scores good at sequencing (=?)

- it is better than structured observation in some cases

Disadvantages of the diary as a method of data collection:

- more expensive

- boring to do, people could stop filling in half way

- takes a long time, people might get less precise along the way

- there is often a time period between when something happens, and when someone fills in their diary

(for example, they do something in the morning, they write it in their diary in the evening) so

memory plays a role


lOMoARcPSD

So that was chapter 9. We’ll continue with the fantastic chapter 10, which is called: ‘Asking

questions’. This chapter is mainly about the way you should or should not state your questions, and

the (dis)advantages of open and closed questions.

Open questions

Advantages:

- respondents can answer in their own terms

- respondents can give answers which are unusual, things the interviewer did not think of before

- the questions do not suggest certain answers

- it is useful for exploring new areas of research

- you can use the answers given to open questions in an interview as a basis to form the questions in a

self-completion questionnaire

Disadvantages:

- time consuming to administer

- coding is needed (= also time consuming)

- require a lot of effort from the respondents

- if there are multiple interviewers, there is the possibility of variability between the interviewers

Closed questions

Advantages:

- easy to process answers

- the answers can be compared better

- sometimes the possible answers that are given, can clarify a question for a respondent

- it is easier to perform for the respondent

- in interviews, closed questions reduce the possibility of variability in the recording of answers in

structured interviewing

Disadvantages:

- a loss of spontaneity in respondents’ answers

- it can be difficult to make forced-choice answers mutually exclusive

- it is hard to let the different answers given as options cover all possible answers

- they can be annoying if you feel like your answer isn’t one of the options given

- difficult to establish rapport (de band die opgebouwd wordt tussen respondent en interviewer

Types of questions

- personal factual questions

- factual questions about others

- informant factual questions

- questions about attitudes

- questions about beliefs

- questions about normative standards and values

- questions about knowledge

-

Rules for designing questions

General rules of thumb:

- always bear in mind your research questions

- what do you want to know?

- how would you answer the question yourself?

Specific rules when designing questions:


lOMoARcPSD


- avoid unambiguous terms in questions

- avoid long questions

- avoid double barreled questions (questions that are about the same thing)

- avoid very general questions

- avoid leading questions (questions in which you steer the respondent towards a certain answer)

- avoid questions that are actually asking two questions (is similar to double barreled I think)

- avoid questions that include negatives

- avoid technical terms

- does the respondent have the requisite knowledge?

- make sure that there is a symmetry between a closed question and its answers

- make sure that the answers provided for a closed question are balanced (e.g.: very bad – bad –

neutral – good – very good)

- memory problems

- don’t know

The last point is about whether or not to add ‘don’t know’ as an option.

Why do: People may else be forced to pick an answer, while they don’t know.

Why do not: People may not take the effort to think about the subject when they see the option

‘don’t know’.

Vignette questions

This technique essentially comprises presenting respondents with one or more scenarios and then

asking them how they would respond when confronted with the circumstances of that scenario.

Piloting and pre-testing questions

Piloting and pre-testing means the same thing: testing your questions before you send them to

everyone. You can do this to find points for improvement, which you haven’t thought of before.

Using existing questions

Advantages:

- they have already been piloted

- if reliability and validity tests have been performed in the research from which you take the

questions, you can say something about the measurement quality

- you can make links to other researches in which these questions have also been asked

Dat was het, lekker dan!!!

Smakelijk ja.


lOMoARcPSD

Samenvatting H12 & H14 gemaakt door: Daan Kuitenbrouwer (dat schrijf je zoals de schrijver)

H12: Content analysis

“Content analysis is an approach to the analysis of documents and texts that seeks to quantify

content in terms of predeterminded categories and in a systematic and replicable manner.”

“Content analysis analysis is any technique for making inferences(conclusions) by objectively and

systematically identifying specified characteristics of messages.”

It is rooted in the quantitative research strategy.

Content analyse is een benadering van texten of documenten(dit is vooral media e.d.). Deze

benadering streeft ernaar alles systematisch op te delen in categorien. Het is daarom ook vooral

nuttig voor kwantitatief onderzoek. Met deze methode kan een (hoofd)onderzoeker hulponderzoekers

als machines texten laten doorspitten, en daaruit in getallen conclusies trekken(coding schedule).

What are research questions?

It is necessary to specify the research question precisely: - media that will be content analysed.

-coding schedule

Typical questions seem to revolve around: who, what, where, location, how much and why.

Selecting a sample.

Content analysis can be applied to many types of documents. Two important distinguishable sample

types are: - different media

-different dates

What is to be counted/taken into account?

This of course depends on your research question. Often used:

-Significant actors -what kind of person has produced the item.

-who is or are the main focus of the item.

-who provides alternative voices.

-what was the context for the item.

-Words -counting of frequency of certain words.

-Subject & Themes -for coding = categorization of the phenomenon of interest(see next section).

-Dispositions(neiging?) -This is for a further level of interpretation of the text:

disposition(neiging/voorkeur) of the producer of the text.

Coding

-Coding schedule -form onto which all data relating to an item being coded will be

entered.(fig12.1Blz:283). Each column is a dimension.

-Coding manual -statement of instructions to coders, that also includes all the possible

categories for each dimension being coded. (fig12.2blz284).

The coding schedule is then filled in, and that is done for more news items. This gives the data for the

study. Check deze beide pagina’s echt, dit verduidelijkt het een hele hoop.

Dit stukje is minder boeiend/belangrijk:

Potential pitfalls(valkuilen) in devising coding schemes(waar de (hoofd)onderzoeker op moet letten):

-Discrete dimensions: make sure no overlap

-mutually exclusive categories: no overlap between categories

-exhaustive categories: For each dimension all possible categories should be available to coders.

-Clear instructions

-Clarity about the unit of analysis.


lOMoARcPSD


Vanaf hier is het wel weer boeiend:

Advantages of content analysis

-Transparent � replications and follow-up studies are feasible.

-Longitudinal analysis(over time) with relative ease.

-No reactive effect(research does not influence texts/documents in comparison: Interviewing does

influence to respondents.)

-Highly flexible method: Applied to different kinds of unstructured information.

-Allows information to be generated about social groups to which it is difficult to gain acces.

Disadvantages of content analysis.

- A content analysis can only be as good as the document it works with.

-Coding manuals cannot be devised so that no interpretation is needed by the

coders(hulponderzoekers)

-Particular problems when the aim is to imput latent(verborgen)content.

-It is difficult to ascertain the answers to ‘why’questions.

-Sometimes these studies are accused of being atheoretical: accent being placed on what is

measurable rather than what is theoretically significant or important.

Samenvatting H14: Quantitative data analysis. Gemaakt door: Daan

One should know what kind of data analysis is going to be used before designing the questionnaire/

observation-schedule / coding frame. Because not every technique is appropriate for any variable.

Types of variables

-interval/Ratio variables: Variables where the distances between the categories are identical across

the range of categories(example:Minutes: 36 is one more than 35)

-Ordinal variables: Variables whose categories can be rank ordered(like interval/ratio) but the

distances between the categories are not equal across the range(example: diff 4-6 days a week –

every day is not the same as diff 4-6 days a week – 2-3 days a week)

-Nominal/categorical variables: comprise categories that cannot be rank ordered.(example:

relaxation is not more of something than lose weight.)

-Dichotomous variables: Contain only two categories(example: gender).

Dit staat allemaal netjes in table 14.1 op pagina 322. Ook fig 14.1 op pag 323 is erg nuttig.

Univariate analysis

Frequency tables: table with amount of something and percentage of total: table 14.2 page 323.

Diagrams(pag.324): interval/ration variable � histogram

for other data types � Pie chart & Bar chart

Measures of central tendency(verschillende methodes van middeling)

-arithmetic mean: total divided by counts. (gemiddelde)

-Median: the mid-point in a distribution(mediaan)

-Mode: value that occurs most frequently

Measures of dispersion(amount of variation in a sample) � standard deviation(standaard afwijking)

All these tables and data are produced by SPSS

Bivariate analysis

Bivariate analysis is concerned with the analysis of two variables at a time in order to uncover

whether the two variables are related. Because different types of variables(see above) have to be

compared, there are several techniques. Fig 14.5(326) shows what techniques for what kind of

variables are appropriate. All these methods recover relationships and not causality. So one cannot


lOMoARcPSD

infer that one variable causes another. Different techniques:

-Contingency table: two frequency tables beside each other, with percentages, so that two variables

can be compared(14.4, pag. 327)

-Pearson’s r: -this is a coefficient between -1 and 1.

-the closer to absolute 1, the stronger the relationship.

-If it is (close to) zero, there is no relationship.

-it is often plotted in a scatter diagram(fig 14.6-9, pag328).

-If the sign is negative, there is a negative relationship.

-Coefficient of determination: expresses how much of the variation in one variable is

due to the other variable. This is done by SPSS.(not really important)

-Spearmans’s rho(ρ). -exactly the same in outcome of calculation as the Pearson’s r.

-used for comparing ordinal variables with other ordinal variables or with

interval/ratio variables. This is because the way of counting of ordinal variables is different than the

way of counting of other ordinal variables and interval/ration variables. (calculated by SPSS).

-Phi(φ) and Cramér’s V -Phi is like Pearsons’s r, but for two dichotomous variables(only two

possibilities)

- Cramér’s V is like Pearsons’s r, but for two nominal variables(variables not

rank ordered). It is always positive.

Multivariate analysis

Multivariate analysis entails the simultaneous analysis of three or more variables, and seeks for

relationships. One should be cautious that such a relationship is not spurious, that is: when each

variable is itself related to a third variable. (fig 14.10, pag: 331, example: frequency of shaving related

to heart attack. Both are related to a lifestyle, but not to each other.)

-Intervening variable: Missing link between two related variables.(example: why is market

orientation related to organizational performance?

market orientation � employee attitudes � organizational performance

-a third variable moderating a relationship: Whether a relationship between two variables holds

for a, but not for b. (example: does a relationship

hold for men, but not for women?) Contingency

tables are useful for investigating this.

Statistical significance

One cannot know whether findings are generalizable to the whole population. Although probability

sampling procedures have been followed, a sampling error(difference between the population and

the sample that you have selected) can always occur. To provide an indication of how confident you

can be in your findings(relations), thank god, there is statistical significance!

-first test : -null hypothesis(there is a/no relationship between a and b.)

-Establish the level of statistical significance (p), that is a measure of the degree of

risk that you might wrongly reject the null hypothesis. Usually p < 0.05.

-Determine the statistical significance of your findings(SPSS)

-Reject or confirm the null hypothesis.

-There are two types of errors in this case type 1: rejecting null hypothesis when it

should be confirmed.(due to sample)

type 2: confirming null hypothesis when it

should be rejected. (due to sample)


lOMoARcPSD


-Chi-square test: It is applied to contingency tables, it allows us to establish how confident we

can be that there is a relationship between two variables.(best wiskundig,

dus wordt bij practicum behandelt, denk ik.)

-Correlation and statistical significance: provides information about the likelihood that the coefficient

will be found in the population of the sample that was taken.

(ook wiskundig verhaal, niet voor deze samenvatting)

-Comparing means and statistical significance:(ook wiskundig verhaaltje)

Hoofdstuk 15 zou Anton doen, maar het schijnt alleen over SPSS te gaan, en is daarom buiten

beschouwing gelaten.

Hfst 16 De 4 traditions van kwalitatief onderzoek:

1. Naturlisme, zoekt de sociale realiteit in zijn eigen termen, zoals het echt is. (rijkelijke beschrijving van

mensen en interacties

2. Ethnomethodology, begrijpen hoe sociale orde is ontstaan, door praten en interacties.

3. Emotionalism, een belang weergeven met subjectiviteit en daardoor toegang krijgen tot de “inside

experience”, dus het innerlijk van mensen

4. Postmodernism, dit is gewoon ‘method talk’, maar dan wel per sociale groep verschillend.

De tijdslijn van kwalitatief onderzoek (9 momenten):

1. The traditional period, 1920-1940, het begon met diepte onderzoek naar slices of life van sommige

bevolkingsgroepen (de strangers), pure positivisme.

2. Modernist phase, 1945-1970, voortborduren op de eerste phase, echter met strengere regels voor

kwalitatief onderzoek, nog steeds lichtelijk positivisme

3. Blurred genres, 1970-1986, periode met epistemological en ontological onderzoek, evenals

theoretische ideeën, die via kwalitatief onderzoek worden onderzocht. Nog steeds lichtelijk

positivisme, maar ze beginnen in te zien dat dit soort data door iedereen anders geïnterpreteerd kan

worden.

4. Crisis of representation, mid 1980’s, een periode waarin onderzoekers nog meer begrijpen dat

kwalitatief onderzoek hun manier is van de realiteit afspiegelen(hun interpretatie) en dat de social

locations ook invloed hebben. Ze begrijpen dat deze onderzoeken gelimiteerde scientific value

hebben

5. Postmodern period of experimental ethnographic writing, mid-1990’s, periode waarin het erg

verschilde hoe ze hun participanten in het onderzoek vermeldden.

6. Post-experimental enquiry, 1995-2000, gaat over 1 publisher maar namelijk AltaMira Press, die het

heeft over experimental en interdisciplinary writing.

7. The methodologically contested present, 2000-2004, grote onenigheid over hoe kwalitatief

onderzoek gezien moet worden en in met welke regels het gedaan moet worden. (gaat eigenlijk nog

steeds door)

8. Now, 2005-nu, het ter discussie stellen van kwalitatief onderzoek, grote waarde bevestigd aan

traditionel onderzoek.

9. The fractured future, Randomized field travels zullen de ene groep bezig houden, de andere groep

met sociale en culurele responsieve studies zal bezig houden.

In kwalitatief onderzoek zijn er meerdere methodes te onderscheiden namelijk,

1. Ethnography/participant observation, het observeren in een sociale omgeving

2. Kwalitatieve interviews

3. Focus groups (key concept 8.2)

4. Het analyseren van gesprekken en toespraken e.d.

5. Het analyseren van teksten en documenten


lOMoARcPSD

Vaak worden meerdere methodes tegelijkertijd gebruikt, en de invloed van theorie is vrij groot in

kwalitatief onderzoek.

The main steps in kwalitatief onderzoek:

1. General research questions, je beginpunt is dit.

2. Selection of relevant site(s) and subjects, hier kijk je waar het kwalitatief onderzoek tot zijn recht

komt

3. Collection of relevant data,

4. Interpretation of data,

5. Conceptual and theoretical work, het samenvoegen van de bevinden, labelen,

5a. Tighter specification of the research question

5b. Collection of further data

6. Writing up findings/conclusion

Dit laat zien date r een belangrijke link is tussen theorie en onderzoek.

Veel kwalitatieve onderzoekers denken dat theorie uit het onderzoek verschijnt, dus nieuwe theorie,

echter hoeft dit niet het geval te zijn en wordt er juist alleen maar (door bijvoorbeeld stap 5a en 5b)

geschoven in al bestaande theorieën.

Concepten in kwalitatief onderzoek

Concepten moeten in, sociaal onderzoek, volgens Blumer niet definitef zijn, niet vast staan. Dit

omdat concepten in elke situatie anders kunnen zijn, daarom moeten concepten volgens hem

sensitizing concepts zijn. Ze moeten een gids als het ware zijn. Ze moeten een algemeen beeld

schetsen. Hierop is kritiek omdat een te algemeen concept niet meer bruikbaar is in onderzoek.

(staat voorbeeldje over ernst concepten op pg 374)

Sampling in kwalitatief onderzoek

Voornamelijk in purposive sampling (hfst 17.4), door de selectie van kennis is dit een goede manier

voor het onderzoek. Ook probability sampling kan worden toegepast(interviews). Dit is handig als je

het naar een grotere populatie wilt vertalen(generaliseren), of als de onderzoeker niet verwacht dat

er bepaalde categorieën van mensen gesampled moeten worden. Bij purposive sampling wordt er

dan vaak op meerdere nivo’s gesampled.

Reliability en validity in kwalitatief onderzoek

Dit is een lastig puntje in kwalitatief onderzoek, want validiteit van de metingen brengt ook

bijbetekenissen van de metingen op, dit punt is dus niet al te belangrijk.

Daarom zijn de begrippen opgesplitst:

- External reliability, in hoeverre kan dit onderzoek worden nagedaan? Lastig te meten, maar door

een sociale rol(als in rollenspel) als onderzoek aan te nemen, wordt het al makkelijker.

- Internal reliability, hoeveel mensen werken eraan mee (onderzoekers, niet onderzochte mensen)

- Internal validity, is er een overeenkomst tussen de verschillende metingen tussen de onderzoekers?

Dit wordt door sommige als een belangrijk criteria gezien.

- External validity, kan dit gegeneraliseerd worden?

Er zijn ook twee alternatieve criteria:

- Trustworthiness, dit bestaat weer uit 4 delen:

o Credibility, geloofwaardigheid (internal validity)

o Transferbility, (external validity)

o Dependability, (reliability)

o Confirmability, objectivity� heeft de onderzoeker met de juiste intenties gehandeld

- Authenticity

o Fairness, altijd zelfde viewpoint gebruikt?

o Ontological authenticity, wordt er daadwerkelijk gezocht naar het beter begrijpen van sociaal milieu


lOMoARcPSD


o Educative authenticity, heeft het educatieve waarde voor de ene groep om de andere groep beter te

begrijpen?

o Catalytic authenticity, heeft het onderzoek beweegredenen gegeven om actie te ondernemen, om

de omstandigheden te veranderen.

o Tacitical authenticity, heeft de onderzoeker de onderzochte mensen genoeg stappen gegeven om

actie te ondernemen?

(laatste 2 horen voornamelijk bij action research(als onderzoek bedoeld is om echt verbeteringen te

brengen voor bepaalde groepen ofzo), zijn niet van levensbelang)

De nadruk is dus meer op kwaliteit van het onderzoek gekomen, daar zijn deze criteria voor, zo zijn er

nog wel meer zoals:

Sensitivity to context, commitment and rigour, transparency and coherence en impact and

importance. (criteria van Hammersley)

Vrij logisch klinkende criteria!

Wat opvalt is dat er eigenlijk niet maar 1 goed criterium is waaraan al het kwalitatief onderzoek moet

voldoen, veel onderzoekers hebben hun eigen meningen.

Key concepts (even tussendoor)

Respondent validation, het delen van de gevonden informatie met de onderzochte mensen.

Hier zijn weer meerdere vormen van:

- Elke respondent individueel informeren

- Per groep of organisatie informeren.

- Per groep of organisatie alleen de bevindingen voor die groep of organisasties geven

Doel hiervan is om te kijken of de metingen, bevindingen, kloppen met wat de respondenten denken.

En dan juist de slechte punten vinden en de redenen hierachter. Echter kan een respondent zich dan

weer anders gedragen of het niet eens bevestigen, omdat ze het vanuit een ander punt zien

bijvoorbeeld.

Triangulation

Het gebruik van meer dan een methode of source of data in het onderzoek. Dan krijg je een soort

mixed methods research

Er zijn enkele vooroordelen over kwalitatief onderzoek:

- Het zien door de ogen van de mensen die bestudeerd worden is een haast onmogelijke taak. Daarom

wordt in het onderzoek wel zoveel mogelijk rekening gehouden met dingen door de ogen van de

mensen die bestudeerd worden. (logischerwijs kunnen hierin fouten begaan worden)

- Het beschrijven van het onderzoek. Kwalitatieve onderzoekers willen veelal hun onderzoek tot in het

diepste detail beschrijven omdat alles relevant kan zijn. Dit kan weleens te ver gaan.

- De nadruk op het proces, kwalitatieve onderzoekers leggen vaak de nadruk op een proces. Hoe iets

veranderd en dergelijke en ze zijn daar vaak jaren mee bezig(met het onderzoek), dit maken ze dus

ook allemaal zelf mee. Dit kan bijvoorbeeld ook worden vekregen in een semi structured of

unstructured interview, waarbij de geïnterviewde als het ware een proces vertelt.

- De flexibiliteit en gelimiteerde structuur van het onderzoek. Dit is omdat een kwalitatief onderzoek

juist het proces wilt onderzoeken, vanuit de ogen van de onderzochte mensen. Hierbij moeten soms

aspecten van het onderzoek kunnen worden aangepast, dus is flexibiliteit en gelimiteerde structuur

nodig.

- Concepten en theorie komen vaak voort uit de verkregen data.

Kritiek op kwalitatief onderzoek:

- Het is te subjectief, de interpretatie is persoonlijk en de kijk op dingen ligt ook aan de onderzoeker.


lOMoARcPSD

- Het is moeilijk na te doen. Dit is sowieso moeilijk in social sciences, omdat het vrij ongestructureerd

is. Maar in kwalitatief onderzoek brengt het nog meer problemen met zich mee, omdat de focus kan

verschillen, de respondenten hun visie enz.

- Het probleem van generaliseren. Er worden te weinig respondenten gebruikt om het te kunnen

generaliseren. (daarom noemen ze het moderatum generalization)

- Lack of transparency, het is lastig na te gaan wat de onderzoeker precies allemaal heeft gedaan.

Enkele tegenstellingen tussen kwantitatief en kwalitatief onderzoek:

Kwantitatief Kwalitatief Numbers Words Point of view researcher Point of view participant Researcher is distant Researcher is close Theory and concepts tested in research Theory and concepts emergent from data Static Process Structured Unstructured Generalization Contextual understanding Hard, reliable data Rich, deep data Macro Micro Behaviour Meaning Artificial settings Natural settings

Deze verschillen zijn niet altijd even duidelijk aan te wijzen, in survey research bijvoorbeeld neigt

kwalitatief al meer naar kwantitatief.

Logischerwijs zijn er ook overeenkomsten

- Beide proberen de gevonden data eerst te reduceren tot het nuttige.

- Beide willen de hoofdvraag beantwoorden

- Data analyse moet gerelateerd worden aan research literature

- Beide vinden variatie belangrijk

- Beide zien frequentie als een opstap naar analyse

- Beide willen dat opzettelijke verdraaiing van de resultaten niet voorkomt

- Beide zijn voor transparantie, iedereen moet zien wat ze gedaan hebben

- Er wordt rekening gehouden met fouten

- Research methodes moeten passend zijn bij de hoofdvraag

Mocht je nog wat willen lezen over feminisme en kwalitatief onderzoek, lijkt mij overbodig, zie

pagina 396. Tja, vrouwen. Dit is gemaakt door degene met een langzaam ontdooiende relatie met

onze geliefde studievereniging.

Samenvatting H18 OV1 door Guido

Interviewing in qualitative research.

• Difference between structured interviewing and qualitative interviewing.

Qualitative approach is: less structured, more interest on interviewee’s point of view, interviewers

can easily depart from there schedule, the interviewer adapts his questions to the answers of the

interviewee, the interviewee is only interviewed once, the interviewer wants rich and detailed

answers instead of structured answers that can be coded easily.

For structured interviews it is all the tegenovergestelde.

• Main characteristics of and differences between unstructured and semi-structured interviewing.

Unstructured interview; interviewer only prepares one question and improvises his other questions.

Semi-structured interview; researcher has a list of questions/topics to be an covered (interview

guide) but can also improvise a bit. When the researcher starts the investigation with a fairly clear


lOMoARcPSD


focus, or more then one person has to do fieldwork, or you are doing a multiple case study, you

mostly choose the semi-structured interview.

• How to devise and use an interview guide for semi-structured interviewing

Interview guide=brief list of memory prompts of areas to cover/questions to be asked.

To prepare a interview guide one should order the questions by topic, formulate questions so they

help you to answer your main question, use relevant language for your interviewees, don’t ask

leading questions, ask general information(age,name etc).

Before the interview, make sure: you’re familiar with the setting in which the interviewee works,

have a good recording machine, quiet and private surroundings. After the interview make notes

about: how the interview went, where it took place, feelings about the interview, the surroundings

• Different kinds of question that can be answered in an interview guide

Most interviews contain all 9 of this types of questions:1. introducing q’s: have you ever?

2. follow-up q’s: what do you mean by that?

3. Probing q’s: could you say some more about that

4. specifying q’s: what did you do then?

5. direct q’s: are you gay?

6. indirect q’s: what do most people around here think of his sexual geaardheid?

7. structuring q’s: I would now like to move on to a different topic.

8. silence : so you give the interviewee the opportunity to explain his self.

9. interpreting q’s: Do you mean that that X has to be Y?

Besides, also pictures can be used.

• Importance of recording and transcribing qualitative interviews

Important to hear the way people say something, you can listen to interviews more than once.

• Approaches to sampling in studies using qualitative interviews

Purposive sampling is the best for qualitative interview research, 2 approaches;

1. snowball sampling, used when there is no sampling frame, use contacts between individuals to trace

additional respondents.

2. theoretical sampling, sampling until your categories have achieved theoretical saturation. Difficult to

know when you have theoretical saturation.

• The significance of qualitative interviewing in feminist research.

Face to face interview=feminist method. When a woman interviews a woman there is a high level of

rapport between interviewer an interviewee, high degree of wederkerigheid from the interviewer,

and a non-hierarchical relationship. this is all good for the results. When a man interviews a woman

this is not the case.

• The advantages and disadvantages of qualitative interviewing relative to participant observation.

Advantages of participant observation in comparison to qualitative interviewing:

You can see the results from different perspectives, more reliable. Dialecten and special words/slang

can be understand better through observation. Easier to uncover unexpected topics because

interviewees are used to obsevator. Research is longitudinal, since observatory is present for a longer

time.

Advantages of qualitative interviewing in comparison to participant observation:

Ethical not right to observe, interview is better then. Maybe the presence of the observer would

result in reactive effects. Interviewing is easier sometimes when your research objects are not active

all the time like football hooligans. Sometimes interviewing can allow access to a wider variety of

people and situations.


lOMoARcPSD

Togethersetting of headpiece 22

Qualitative data analysis

Contrary to quantative data analysis, there are very few well-established or widely accepted rules on

how to analyze qualitative data. Most writers would argue that this isn’t even desirable anyway.

There are, however, broad guidelines that can guide you through qualitative data analysis.

General strategies of qualitative data analysis

Two general strategies are discussed

1. Analytic induction

Analytic induction is an approach to the analysis of data in which the researcher seeks universal

explanations of phenomena. He does this by making a hypothetical explanation of a research

question and then examines whether the cases are consistent with his theory. If this is not the case

he reformulates his hypothesis and examines the cases again, or he refines his explanation to exclude

the cases that are not consistent (see figure 22.1)

2. Grounded theory

Grouded theory is the most used framework for analyzing qualitative data analysis. It is best

understood by looking at figure 22.1 on page 545.

Tools of grounded theory:

Theoretical sampling – the process of data collection for generating theory whereby the analyst

collects, codes and analyzes his data jointly and then decides what data to collect next in order to

develop his theory as it emerges.

Coding - The key process of grounded theory is coding. Coding is done by breaking down data into

component parts which are given names. It is advisable to start coding soon after the collection of

data.

Theoretical saturation – Sampling is carried on with untill no new data emerges.

Constant comparison – A process of maintaining a close connection between concepts and

categories to compare phenomena being coded. This is done so theoretical elaborations can emerge.

An important aid in grounded theory is the memo. A note made for the researcher or his collegaus to

remind them about the terms being used and what they mean.

Very similar to Grounded theory is Thematic analysis. Thematic Analysis is an approach to dealing

with data that involves the creation and application of ‘codes’ to data. The data being analysed might

take any number of forms – an interview transcript, field notes, policy documents, photographs,

video footage. As I said before, there is a clear link between this type of analysis and Grounded

Theory, as the latter clearly lays out a framework for carrying out this type of code-related analysis.

Narrative analysis is an approach to analyzing data in the form of a story of some event. It’s purpose

is not to discover whether the story told is true, but to uncover the way the story-teller makes sense

of the event.

The last part of chapter 22 is about secondary analysis (chapter 13) and that this is done more lately

than in the past.


lOMoARcPSD


Recap chapter 24:

The distinction between quantitative and qualitative has been employed so far for two main reasons:

• There are differences between quantitative and qualitative research in terms of research strategy, and many

researchers and writers on research methodology perceive this to be the case.

• It is a useful means of organizing research methods and approaches to data analyze.

Quantitative research teaches us that it is rarely the case that we find perfect associations between variables.

The natural science model and qualitative research:

Qualitative interviews may often reveal a predisposition towards or a reflection of an interpretive and

constructionist position. Although this is not always the case.

Further, qualitative research frequently exhibits features that one would associate with a natural science

model. This tendency is revealed in several ways:

• Empiricist overtones. The empiricism of qualitative research is most notable in conversation analysis. This is an

approach that takes precise transcriptions of talk as it starting point and applies rules of analysis to such data.

• A specific problem focus. Qualitative research can be employed to investigate quite specific, tightly defined

research questions of the kind normally associated with natural science model of the research process.

• Hypothesis- and theory-testing. Qualitative researchers typically discuss hypothesis- and theory-testing in

connection with hypotheses or theories generated in the course of conducting research, as in analytic

induction or grounded theory

• Realism. Realism is one way in which the epistemological basis of the natural sciences has been construed.

Writers on qualitative research sometimes distinguish stances on qualitative research that contain elements of

both qualitative and quantitative research.

Quantitative research and interpretivism:

The widespread inclusion of questions about attitudes in social surveys suggests that quantitative researchers

are interested in matters of meaning. It might be objected that survey questions do not really tap issues of

meaning because they are based on categories devised by the designers of the interview schedule or

questioner. Two points are relevant:

• Absence of respondent validation exercises, the notion that qualitative research is more adept at gaining

access to the point of view of those being studied than quantitative research is invariably assumed rather than

demonstrated.

• If the design of attitude questions is based on prior questioning that seeks to bring out the range of possible

attitudinal positions on an issue attitudinal questions may be better able to gain access to meaning.

Quantitative research and construction

Qualitative content analysis has played an important role in developing just such an understanding, just as

discourse analysis has in relation to the social construction of events and meanings in newspaper reports and tv

programmes. Quantitative research can play a significant role in relation to a constructionist stance.

Research methods and epistemological and ontological considerations.

There are differences between qualitative and quantitative research in terms of their epistemological and

ontological commitments, but the connection between research strategy, on the one hand, and

epistemological and ontological on the other hand is not determistic. There is a tendency for qualitative and

quantitative research to be associated with the epistemological and ontological positions, but the connections

are not perfect.

Problems with the quantitative/qualitative contrast.

Chapter 16 drawn contrasts between qualitative and quantitative research. But there is a risk that this kind of

representation tends to exaggerate the differences between them.

A few of the distinctions will be examined to demonstrate this point.

• Behaviour versus Meaning. The distinction is sometimes drawn between a focus on behaviour and a focus on

meanings. However, quantitative research frequently involves the study of meanings in the form of attitude

scales and other techniques

• Theory and concepts tested in research vs. theory and concepts emergent from data. A further related pointy

s that the suggestion that theory and concepts are developed prior to undertaking a study in quantitative

research is something of a caricature that is true only up to a point.


lOMoARcPSD

• Numbers versus words.

• Artificial versus natural. qualitative research is often viewed as naturalistic. But when qualitative research is

based on interviews, the depiction ‘natural’ is possibly less applicable.

The mutual analysis of quantitative and qualitative research.

On further way in which the barriers between quantitative and qualitative research might be undermined is by

virtue of developments in which each is used as an approach to analyze the other.

• A qualitative research approach to quantitative research. Quantitative research reveals a concern in both

cases with the notion that the written of research not only constitutes the presentation of findings but is also

an attempt to persuade the reader of the credibility of those findings.

• A quantitative research approach to qualitative research. One approach to synthesing related qualitative

studies is meta-ethnography, which is a qualitative research approach to such aggregation.

Quantification in qualitative research

Three observations are worth making about quantification in the analysis and writing-up of qualitative data.

• Thematic analysis.One of the commonest approaches to qualitative data analysis is undertaking a search for

themes in transcript or field notes.

• Quasi-quantification in qualitative research. Qualitative researchers engage in ‘quasi0quantification’through

the use of terms such as ‘may’, ‘frequently’, ‘rarely’, ‘often’ and ‘some’. But the researcher should have some

idea of the relative frequency of the phenomena being referred to.

• Combating anecdotalism through limited quantification. A critic point that is levelled against q qualitative

research is that the publications on which it is based are often anecdotal, giving the reader little guidance as to

the prevalence of the issue to which the anecdote refers. A response to this problem is that qualitative

researchers sometimes undertake a limited amount of qualification of their data.

Summary chapter 25: Benno L.

Mixed methods research: combining quantitative and qualitative research

Definition

Mixed strategy research is a term used this book to describe investigations combining quantitative

and qualitative research. But mixed methods research has become the preferred term, because it

emphasizes the mixing of the research methods and not just using them in tandem. In other words:

the quantitative and qualitative data derived from mixed methods research should be ‘mutually

illuminating’ (mijn vertaling: wederzijds aanvullend/verklarend)

(Kijk uit dat Robin niet ineens in je bed op duikt, want als ze er ligt komt ze niet meer van je af, ik kan

het weten, ik ben Benno L.)

Argument against mixed methods research

In het boek vind ik het niet heel helder uitgelegd, dus volgende stukje is vanuit het book

gecombineerd met een artikel:

The argument is based on two ways of thinking:

1. The idea that research methods carry epistemological commitments, and

2. The idea that quantitative and qualitative research are separate paradigms.

Recap: Epistemological issue concerns the question of what is (or should be) regarded as acceptable

knowledge in a discipline

The first argument comes from the idea that every method is embedded in epistemological and

ontological commitments. A couple of criticisms are: The two methods are not complementary

because they have totally different epistemological implications. And the integration of research

strategies ignores the assumptions underlying research methods and transforms qualitative research


lOMoARcPSD


into a procedural variation of quantitative research.

This leads to the main difficulty: The fixed epistemological and ontological implications from both

methods are very difficult to sustain.

(Eigenschappen en aannames van beide methodes worden gemengd).

Paradigm: a cluster of beliefs and dictates which for scientists in a particular discipline influence what

should be studied, how research is done and how results should be interpreted (Brymann). Important

of paradigms is that they are incommensurable, they are inconsistent with each other because of

their divergent assumptions and methods.- ALLEJEZUS WAT EEN KUT BEGRIP

Komt ie: Een soort gedachte patroon in a scientific discipline die men volgt bij het doen van

onderzoek. Paradigms zijn incommensurable (niet te vergelijken met andere paradigma’s).

Second argument is closely related and includes paradigms. Quantitative and qualitative research are

seen as paradigms, they are built on fixed way of thought patterns. Or in other words, according to

the paradigmatic position, qualitative and quantitative research are seen to be intrinsically different

beasts underpinned by different philosophical assumptions. Zoek op blz 604/Wikipedia als je nu nog

niet snapt.

Two versions of the debate:

• Epistemological version

incompatible epistemological principles of quantitative and qualitative research

e.g. embedded methods /paradigm arguments

• Technical/pragmatic version

quantitative and qualitative research can be combined

relative strengths and weaknesses of each for data collection / analysis

Approaches to mixed methods research

Couple of reseachers have proposed different approaches. The most important one’s:

Hammersley (1996)

– triangulation – quantitative to confirm qualitative of vice versa

– facilitation - using one research strategy to facilitate the other

– complementarity - Combining the best aspects of two research strategies

Morgan (1998)

– Priority decision - how far is a quantitative or qualitative method the principal data-gathering tool?

– Sequence decision - which method precedes which?

There is a list of 18 different approaches in Bryman’s book. If you want to know them all, look at page

608/609 and later on page 611 til 623.

Common differences

Static and processual features

quantitative research uncovers regularities

qualitative research reveals social processes

Researchers’ and participants’ perspectives

quantitative methods to test researcher’s theories

qualitative methods to discover actors’ meanings


lOMoARcPSD

When successful?

Success depends on 4 factors:

well designed and conducted

methods appropriate to research questions

effects of spreading limited resources

skills and training of researchers

Not inherently superior to mono-method or mono-strategy research

Chapter 27 Gemaakt door de man die in staat is Neutronen te splijt: Jimmy Neutron!!!

Witing up a social research

Many people find writing up a research more difficult than carrying it out. Your readers have to be

convinced about the credibility of the knowledge claims you are making. Good writing is persuasive

and convincing.

Key concept 27.1 What is rhetoric? The way in which attempts to convince or persuade an audience

are formulated.

- The first bit of advice is Start early with writing up your research.

- You also should be persuasive. Convince your readers of the credibility of your conclusions.

- Get feedback. Get as much feedback and response as possible.

- Avoid sexist, racist, and disablist language.

Structure you’re writing.

- Title page

- Acknowledgements (pay attention to the ones how helped you and gave you feedback)

- List of contents

- An abstract (brief summary)

- Introduction. - Explain what you are writing about and why it is important.

- Indicate the general theoretical approach you will be using

- Outline your research questions

- A good opening sentence which is not to evasive (niet te open)

- Literature review (zie hoofdstuk 4)

- Research methods. – Research design (the procedures you used)

- Nature of your questionnaire, interview schedule, population

- Results - Do not include ALL results. Only the results that relate to the research questions

- Point to particular parts of your results. Do not just summarize what a table shows.

- Also describe the results don’t just present a graph or a table without any comment.

- When reporting quantitative findings, try to vary the results if possible (variation)

- The major problem we face in qualitative inquiry is not to get data, but to get rid of it.

- If you have more chapters with results, than show in the beginning of the chapter which research

questions will be addressed and summarize it at the end of the chapter.

- Discussion. How do the results illuminate your research questions? Are the hypotheses confirmed or

not. And speculate what could be improved.

- Conclusion. - Relating you findings and discussion to your research questions.

- Make clear the implications of our findings for your research questions.


lOMoARcPSD


- Suggest how the findings influence theories

- Draw attention to limitations of your research. (critics, improvements)

- Propose areas of further research

- Don’t mention NEW things. And don’t speculate too far away from the data.

- Appendices. Questionnaire, coding frame, observation schedule.

- References. All references cited in the text.

Writing up quantitative, qualitative and mixed methods research.

Quantitative researchers often give more detailed accounts of their research design, research

methods, and approaches to analysis than qualitative researchers. However we should not be too

surprised if they turn out to be more similar than might have been expected.

Writing up quantitative research

If you want your article published. Notice that 90% of the articles for prestigious publications are

rejected. Also they can give you feedback to change certain things before it is excepted.

Structure: - Abstract

- Introduction. Try to grab the attention of the audience in the first sentences.

- Theory. Hypothesis.

- Data

- Measurment. Concepts and scale

- Methods and models. relationships

- results

- conclusion

Good points about Kelley and De Graaf’s article.

- Clear attempt to grap the readers attention.

- Spell out clearly the rationale of their research.

- Research questions are spelled out very specific.

- Clear and explicitly summarized measurement of concepts, sampling, research methods and

approaches.

- The findings presented are very specific tot the research questions.

- The conclusion returns to the research questions and give implications.

Writing up qualitative research

Now a study of vegetarianism is looked at:

Structure: - Abstract

- Introduction. Give immediate sense of what the article is about.

- The analysis of the social dimensions. They say not many social scientists have paid

attention to vegetarianism yet.

- other studies

- The design of the study

- Explaining

- Conclusions

What are the lessons:

- Strong opening sentences that attract our attention an give a clear view of the article.

- The rationale of the research is clearly identified

- Research questions are specified but they are more open then Kelley and De Graaf’s.


lOMoARcPSD

27.2 Postmodernism and its implications for writing

What is postmodernism?

Postmodernism as an approach is at least two things. One is an attempt to get grips with the nature

of modern society and culture. The other is that it represents a way of thinking about and

representing the nature of the social schiences and their claims to knowledge.

Wikipedia: Het meest in het oog springende kenmerk van de stroming is het in twijfel trekken van

lang gekoesterde begrippen als waarheid en - romantische - authenticiteit.

27.3 What is the linguistic turn?

The linguistic turn is based on the idea that language shapes our understanding of the world.

27.4 What is reflexivity?

The way in which speech and action are constituitive of the social world in which tey are located.

In other words, they do more than merely act as indicators of deeper phenomena.

Writing ethnography

Zie hoofdstuk 17.

Refers both to a method of social research and to the finished product of ethnographic research.

Wikipedia over etnografie: betekent letterlijk het in kaart brengen van etnieën. In de sociolinguïstiek

duidt het woord op een systematische methode om, door bestudering van plaatselijke attitudes, en

niet het minst talige attitudes, verschillende bevolkingsgroepen te onderscheiden.

27.5 Three forms of ethnographic writing.

1. Realist tales.

2. Confessional tales

3. Impressionist tales.

Experiential authority

The author provides a narrative in which he or she is no longer to be seen. As a result, an impression

is conveyed that the findings presented are what any reasonable, similarly placed researcher would

have found.

Typical forms

The author is generalizing about a number of recurring features of the group in question to create a

typical form that that feature takes.

The native’s point of view

The commitment to seeing through the eyes of the people being studied. This is an important feature

for qualitative researchers because it is part of a strategy of getting at the meaning of social reality

from the perspective of those being studied.

Interpretative omnipotence

The author rarely presents possible alternative interpretations. Instead the phenomenon in question

is presented as having a single meaning or significance.

Eindwoord van onze begeleider Anton,


lOMoARcPSD



lOMoARcPSD

Management of innovation processes

Documents

social research

interpretation of research

innovation research

research method

h2 research design

social actors

social phenomena

main research designs